Principal Components Analysis

  • B. Roy Frieden
Chapter
Part of the Springer Series in Information Sciences book series (SSINF, volume 10)

Abstract

A concept that is closely related to linear regression (preceding chapter) is principal components [15.1]. Linear regression addressed the question of how to fit a curve to one set of data, using a minimum number of factors. By contrast, the principal components problem asks how to fit many sets of data with a minimum number of curves. The problem is now of higher dimensionality. Specifically, can each of the data sets be represented as a weighted sum of a “best” set of curves? Each curve is called a “principal component” of the data sets.

Keywords

Principal Component Analysis Sample Variance Sample Covariance Matrix Spectral Window Multispectral Imagery 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in to check access.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 15.1
    H. Hotelling: J. Educ. Psych. 24 ,417, 498 (1933)Google Scholar
  2. 15.2
    J. L. Simonds: J. Phot. Sci. Eng. 2, 205 (1958)Google Scholar
  3. 15.3
    R. C. Gonzalez, P. Wintz: Digital Image Processing (Addison-Wesley, Reading, MA 1977) pp. 310–314MATHGoogle Scholar
  4. 15.4
    M. G. Kendall, A. Stuart: The Advanced Theory of Statistics, Vol. 3 (Charles Griffin, London 1968)Google Scholar
  5. 15.5
    M. A. Girshick: Ann. Math. Statist. 10, 203 (1939)MathSciNetMATHCrossRefGoogle Scholar

Additional Reading

  1. Wilkinson, J. H.: The Algebraic Eigenvalue Problem (Clarendon, Oxford 1965)MATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1991

Authors and Affiliations

  • B. Roy Frieden
    • 1
  1. 1.Optical Sciences CenterThe University of ArizonaTucsonUSA

Personalised recommendations